This doctoral thesis focuses on the evaluation of public policies at the municipal level in Italy, with specific attention to territorial cohesion policies and income support measures. The first chapter analyzes the impact of the funds allocated within the framework of the National Strategy for Inner Areas, a place-based policy aimed at combating depopulation in the most remote and marginal municipalities. Using a Staggered Difference-in-Difference econometric model, the study evaluates the effects of financial support on the demographic structure and the number of economic activities at the municipal level. The results show a significant increase in new business units in the treated municipalities, without notable changes in the population structure. Additionally, the presence of spillover effects on neighboring municipalities is analyzed, revealing a positive impact on the increase in economic activities in the surrounding areas. The second chapter investigates the features of delays in cohesion infrastructure projects in Italy. Using a rich dataset at the project and municipal levels, the analysis predicts the socio-economic and institutional features relevant to project timelines through the use of advanced Machine Learning techniques. The efficiency of educational institutions, administrative capacity, and allocated funds are found to be key factors that significantly influence delays. The study further demonstrates that Machine Learning can improve the efficiency of public investments by overcoming logistical and administrative barriers associated with the implementation of cohesion policies. The third chapter examines the impact of poverty and inequality on the distribution of the Citizenship Income at the municipal level, a minimum income policy for families in economic hardship. Using spatial econometric models, the research explores the correlation between poverty, inequality, and program participation. The results indicate that areas with greater socio-economic difficulties show higher participation, while wealthier regions exhibit lower participation. Furthermore, it is noted that the COVID-19 pandemic has exacerbated territorial disparities, making these dynamics even more pronounced. The fourth chapter explores the relationship between the Citizenship Income and voting behavior, taking into account the socio-spatial differences across Italy. The study combines administrative data on program beneficiaries with municipal-level electoral data. Using a continuous treatment Difference-in-Difference model, the analysis investigates how financial support has influenced voting for the governing party. The results show that in disadvantaged areas with lower institutional quality, the policy has increased electoral support for the governing party, while in more prosperous areas, the opposite effect was observed. In conclusion, the thesis provides an in-depth analysis of public policies. The use of econometric, counterfactual, and spatial models, along with predictive Machine Learning algorithms, supported by granular municipal data, has enabled a detailed examination of socio-economic disparities in Italy, surpassing the traditional analyses based on regional and macro-regional divisions. The research highlights how cohesion and income support policies can be optimized through targeted planning tailored to local specificities. The findings offer useful policy insights for policymakers, providing concrete tools to reduce territorial inequalities and promote more balanced and inclusive development across the country.
La presente tesi di dottorato si concentra sulla valutazione delle politiche pubbliche a livello comunale in Italia, con focus specifici sulle politiche di coesione territoriale e sulle misure di sostegno al reddito. Il primo capitolo analizza l’impatto dei finanziamenti destinati nell’ambito della Strategia Nazionale delle Aree Interne, una politica place-based volta a contrastare lo spopolamento nei comuni più distanti e marginali. Attraverso un modello econometrico Staggered Difference-in-Difference, lo studio valuta gli effetti del supporto finanziario sulla struttura demografica e sul numero di attività economiche a livello comunale. I risultati evidenziano un aumento significativo di nuove unità locali nei comuni trattati, senza cambiamenti rilevanti nella struttura della popolazione. Viene inoltre analizzata la presenza di effetti di spillover sui comuni limitrofi, con un impatto positivo sull’aumento delle attività economiche nelle aree circostanti. Il secondo capitolo indaga le determinanti dei ritardi nei progetti infrastrutturali di coesione in Italia. Avvalendosi di un ricco dataset a livello di progetto e comune, l’analisi predice le caratteristiche socio-economiche e istituzionali rilevanti per le tempistiche progettuali, mediante l’utilizzo di avanzate tecniche di Machine Learning. L’efficienza delle istituzioni educative, la capacità amministrativa e i fondi assegnati risultano essere fattori che influenzano significativamente i ritardi. Lo studio dimostra inoltre che il Machine Learning può migliorare l’efficienza degli investimenti pubblici, superando ostacoli logistici e amministrativi legati all’implementazione delle politiche di coesione. Il terzo capitolo esamina l’impatto di povertà e disuguaglianze sulla distribuzione del Reddito di Cittadinanza a livello comunale, una politica di reddito minimo per famiglie in difficoltà economica. Utilizzando modelli econometrici spaziali, la ricerca esplora la correlazione tra povertà, disuguaglianza e partecipazione al programma. I risultati indicano che le aree con maggiori difficoltà socio-economiche mostrano una più alta adesione, mentre le regioni più prospere una minore partecipazione. Si rileva inoltre come la pandemia da COVID-19 abbia acuito le disparità territoriali, rendendo queste dinamiche più evidenti. Il quarto capitolo esplora la relazione tra Reddito di Cittadinanza e comportamento elettorale, considerando le differenze socio-spaziali italiane. Lo studio combina dati amministrativi sui beneficiari del programma con dati elettorali comunali. Utilizzando un modello di Difference-in-Difference a trattamento continuo viene analizzato come il sostegno economico abbia influenzato il voto per il partito di governo. I risultati mostrano che, nelle aree svantaggiate e con minore qualità istituzionale, la politica ha incrementato il sostegno elettorale per il partito al governo, mentre in aree più prospere si è verificato l’effetto contrario. In conclusione, la tesi offre una visione approfondita delle politiche pubbliche. L’utilizzo di modelli econometrici, controfattuali, spaziali e di algoritmi predittivi di Machine Learning, supportati da dati granulari comunali, ha consentito di analizzare dettagliatamente le disparità socio-economiche italiane, superando le classiche analisi basate su aree regionali e macroregionali. La ricerca evidenzia come le politiche di coesione e di sostegno ai redditi possano essere ottimizzate attraverso una pianificazione mirata alle specificità locali. I risultati forniscono spunti di policy utili per i policy maker, offrendo strumenti concreti per ridurre le disuguaglianze territoriali e promuovere uno sviluppo più equilibrato e inclusivo in tutto il Paese.
Valutazione Granulare delle Politiche: Analisi Socio-Economica e Spaziale in un Paese Disuguale
MONTURANO, GIANLUCA
2025
Abstract
This doctoral thesis focuses on the evaluation of public policies at the municipal level in Italy, with specific attention to territorial cohesion policies and income support measures. The first chapter analyzes the impact of the funds allocated within the framework of the National Strategy for Inner Areas, a place-based policy aimed at combating depopulation in the most remote and marginal municipalities. Using a Staggered Difference-in-Difference econometric model, the study evaluates the effects of financial support on the demographic structure and the number of economic activities at the municipal level. The results show a significant increase in new business units in the treated municipalities, without notable changes in the population structure. Additionally, the presence of spillover effects on neighboring municipalities is analyzed, revealing a positive impact on the increase in economic activities in the surrounding areas. The second chapter investigates the features of delays in cohesion infrastructure projects in Italy. Using a rich dataset at the project and municipal levels, the analysis predicts the socio-economic and institutional features relevant to project timelines through the use of advanced Machine Learning techniques. The efficiency of educational institutions, administrative capacity, and allocated funds are found to be key factors that significantly influence delays. The study further demonstrates that Machine Learning can improve the efficiency of public investments by overcoming logistical and administrative barriers associated with the implementation of cohesion policies. The third chapter examines the impact of poverty and inequality on the distribution of the Citizenship Income at the municipal level, a minimum income policy for families in economic hardship. Using spatial econometric models, the research explores the correlation between poverty, inequality, and program participation. The results indicate that areas with greater socio-economic difficulties show higher participation, while wealthier regions exhibit lower participation. Furthermore, it is noted that the COVID-19 pandemic has exacerbated territorial disparities, making these dynamics even more pronounced. The fourth chapter explores the relationship between the Citizenship Income and voting behavior, taking into account the socio-spatial differences across Italy. The study combines administrative data on program beneficiaries with municipal-level electoral data. Using a continuous treatment Difference-in-Difference model, the analysis investigates how financial support has influenced voting for the governing party. The results show that in disadvantaged areas with lower institutional quality, the policy has increased electoral support for the governing party, while in more prosperous areas, the opposite effect was observed. In conclusion, the thesis provides an in-depth analysis of public policies. The use of econometric, counterfactual, and spatial models, along with predictive Machine Learning algorithms, supported by granular municipal data, has enabled a detailed examination of socio-economic disparities in Italy, surpassing the traditional analyses based on regional and macro-regional divisions. The research highlights how cohesion and income support policies can be optimized through targeted planning tailored to local specificities. The findings offer useful policy insights for policymakers, providing concrete tools to reduce territorial inequalities and promote more balanced and inclusive development across the country.File | Dimensione | Formato | |
---|---|---|---|
PhD_Thesis_Definitive_Monturano.pdf
accesso aperto
Dimensione
17.09 MB
Formato
Adobe PDF
|
17.09 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/196052
URN:NBN:IT:UNIMORE-196052